OBJECTIVES A Bayesian network meta-analysis was performed comparing the efficacy and safety of drug-eluting balloons (DEB), drug-eluting stents (DES), or plain old balloon angioplasty (POBA) for treatment of in-stent restenosis (ISR). BACKGROUND Optimal treatment options for ISR have not been well established. METHODS Randomized, controlled trials comparing DEB, DES, and POBA for the treatment of ISR after percutaneous coronary intervention with bare metal stent or DES were included. The primary outcome was target lesion revascularization (TLR). The pairwise posterior median odds ratio (OR) with 95% credible interval (CrI) was the effect measure. RESULTS This analysis included 2,059 patients from 11 RCTs. The risk of TLR was markedly lower in patients treated with DEB (OR: 0.22, 95% CrI: 0.10 to 0.42) or DES (OR: 0.24, 95% CrI: 0.11 to 0.47) than in those treated with POBA in a random-effects model. In a comparison of DEB and DES, the risk of TLR (OR: 0.92, 95% CrI: 0.43 to 1.90) was similar. The risk of MI or all-cause mortality was lowest in the DEB group compared with the DES and POBA groups, which did not meet statistical significance. The risk of major adverse cardiac events, which was mainly driven by TLR, was also significantly lower in the DEB or and DES group (OR: 0.28, 95% CrI: 0.14 to 0.53) than in the POBA group, but it was similar between the DEB and DES groups (OR: 0.84, 95% CrI: 0.45 to 1.50). The probability of being ranked as the best treatment was 59.9% (DEB), 40.1% (DES), and 0.1% (POBA) in terms of TLR, whereas it was 63.0% (DEB), 35.3% (POBA), and 1.7% (DES) in terms of MI. CONCLUSIONS Local drug delivery by DEB or DES for ISR lesions was markedly better than POBA in preventing TLR, but not for MI or mortality. Among the 2 different strategies of drug delivery for ISR lesions, treatment with DEB showed a trend of less development of MI than did treatment with DES.